path: root/sci-libs/datasets/files/datasets-2.17.1-tests.patch

--- a/tests/test_arrow_dataset.py	2024-02-20 21:53:24.248470991 +0100
+++ b/tests/test_arrow_dataset.py	2024-02-20 21:53:29.441804737 +0100
@@ -4016,7 +4016,6 @@
     [
         "relative/path",
         "/absolute/path",
-        "s3://bucket/relative/path",
         "hdfs://relative/path",
         "hdfs:///absolute/path",
     ],
@@ -4136,6 +4136,7 @@
                 )
                 self.assertDictEqual(features_after_cast, dset.features)
 
+    @pytest.mark.skip(reason="require soundfile")
     def test_task_automatic_speech_recognition(self):
         # Include a dummy extra column `dummy` to test we drop it correctly
         features_before_cast = Features(
--- a/tests/test_load.py	2024-02-20 22:12:13.699209107 +0100
+++ b/tests/test_load.py	2024-02-20 22:13:10.862626708 +0100
@@ -388,6 +388,7 @@
             hf_modules_cache=self.hf_modules_cache,
         )
 
+    @pytest.mark.skip(reason="")
     def test_HubDatasetModuleFactoryWithScript_dont_trust_remote_code(self):
         # "lhoestq/test" has a dataset script
         factory = HubDatasetModuleFactoryWithScript(
@@ -403,6 +404,7 @@
         )
         self.assertRaises(ValueError, factory.get_module)
 
+    @pytest.mark.skip()
     def test_HubDatasetModuleFactoryWithScript_with_github_dataset(self):
         # "wmt_t2t" has additional imports (internal)
         factory = HubDatasetModuleFactoryWithScript(
@@ -412,6 +414,7 @@
         assert importlib.import_module(module_factory_result.module_path) is not None
         assert module_factory_result.builder_kwargs["base_path"].startswith(config.HF_ENDPOINT)
 
+    @pytest.mark.skip()
     def test_GithubMetricModuleFactory_with_internal_import(self):
         # "squad_v2" requires additional imports (internal)
         factory = GithubMetricModuleFactory(
@@ -420,6 +423,7 @@
         module_factory_result = factory.get_module()
         assert importlib.import_module(module_factory_result.module_path) is not None
 
+    @pytest.mark.skip()
     @pytest.mark.filterwarnings("ignore:GithubMetricModuleFactory is deprecated:FutureWarning")
     def test_GithubMetricModuleFactory_with_external_import(self):
         # "bleu" requires additional imports (external from github)
@@ -1033,6 +1037,7 @@
         datasets.load_dataset_builder(SAMPLE_DATASET_TWO_CONFIG_IN_METADATA, "non-existing-config")
 
 
+@pytest.mark.skip()
 @pytest.mark.parametrize("serializer", [pickle, dill])
 def test_load_dataset_builder_with_metadata_configs_pickable(serializer):
     builder = datasets.load_dataset_builder(SAMPLE_DATASET_SINGLE_CONFIG_IN_METADATA)
@@ -1154,6 +1159,7 @@
     assert len(builder.config.data_files["test"]) > 0
 
 
+@pytest.mark.skip()
 def test_load_dataset_builder_fail():
     with pytest.raises(DatasetNotFoundError):
         datasets.load_dataset_builder("blabla")
@@ -1169,6 +1175,7 @@
     assert isinstance(next(iter(dataset["train"])), dict)
 
 
+@pytest.mark.skip()
 def test_load_dataset_cached_local_script(dataset_loading_script_dir, data_dir, caplog):
     dataset = load_dataset(dataset_loading_script_dir, data_dir=data_dir)
     assert isinstance(dataset, DatasetDict)
--- a/tests/test_hf_gcp.py	2024-02-21 09:59:26.918397895 +0100
+++ b/tests/test_hf_gcp.py	2024-02-21 09:59:46.335100597 +0100
@@ -45,6 +45,7 @@
         ]
 
 
+@pytest.mark.skip("network")
 @parameterized.named_parameters(list_datasets_on_hf_gcp_parameters(with_config=True))
 class TestDatasetOnHfGcp(TestCase):
     dataset = None
--- a/tests/test_inspect.py	2024-02-21 10:03:32.315520016 +0100
+++ b/tests/test_inspect.py	2024-02-21 10:03:50.345553490 +0100
@@ -49,6 +49,7 @@
     assert list(info.splits.keys()) == expected_splits
 
 
+@pytest.mark.skip(reason="require network")
 def test_get_dataset_config_info_private(hf_token, hf_private_dataset_repo_txt_data):
     info = get_dataset_config_info(hf_private_dataset_repo_txt_data, config_name="default", token=hf_token)
     assert list(info.splits.keys()) == ["train"]
--- a/tests/test_data_files.py	2024-02-21 20:22:57.536160356 +0100
+++ b/tests/test_data_files.py	2024-02-21 20:25:00.153052174 +0100
@@ -378,6 +378,7 @@
         assert len(hub_dataset_repo_patterns_results[pattern]) == 0
 
 
+@pytest.mark.skip(reason="network")
 def test_DataFilesList_from_patterns_locally_with_extra_files(complex_data_dir, text_file):
     data_files_list = DataFilesList.from_patterns([_TEST_URL, text_file.as_posix()], complex_data_dir)
     assert list(data_files_list) == [_TEST_URL, text_file.as_posix()]
@@ -467,6 +468,7 @@
         assert Hasher.hash(data_files1) != Hasher.hash(data_files2)
 
 
+@pytest.mark.skip(reason="network")
 def test_DataFilesDict_from_patterns_locally_or_remote_hashing(text_file):
     patterns = {"train": [_TEST_URL], "test": [str(text_file)]}
     data_files1 = DataFilesDict.from_patterns(patterns)
--- a/tests/packaged_modules/test_folder_based_builder.py	2024-02-21 21:30:20.718922523 +0100
+++ b/tests/packaged_modules/test_folder_based_builder.py	2024-02-21 21:31:46.309061287 +0100
@@ -382,6 +382,7 @@
         assert example[column] is not None
 
 
+@pytest.mark.skip(reason="network")
 @pytest.mark.parametrize("remote", [True, False])
 @pytest.mark.parametrize("drop_labels", [None, True, False])
 def test_data_files_with_different_levels_no_metadata(
@@ -405,6 +406,7 @@
         assert all(example.keys() == {"base", "label"} for _, example in generator)
 
 
+@pytest.mark.skip(reason="network")
 @pytest.mark.parametrize("remote", [False, True])
 @pytest.mark.parametrize("drop_labels", [None, True, False])
 def test_data_files_with_one_label_no_metadata(data_files_with_one_label_no_metadata, drop_labels, remote, cache_dir):
--- a/tests/test_metric_common.py	2023-05-04 18:48:48.550861318 +0200
+++ b/tests/test_metric_common.py	2023-05-04 18:50:25.787364577 +0200
@@ -93,6 +93,7 @@
     INTENSIVE_CALLS_PATCHER = {}
     metric_name = None
 
+    @pytest.mark.skip(reason="disabling, depends on bert_score, bleurt, math_equivalence, coval, nltk, faiss, mauve, rouge_score, sacrebleu, sacremoses ...")
     @pytest.mark.filterwarnings("ignore:metric_module_factory is deprecated:FutureWarning")
     @pytest.mark.filterwarnings("ignore:load_metric is deprecated:FutureWarning")
     def test_load_metric(self, metric_name):
--- a/tests/test_distributed.py	2023-05-04 19:43:09.861275030 +0200
+++ b/tests/test_distributed.py	2023-05-04 19:44:17.608326722 +0200
@@ -74,6 +74,7 @@
         split_dataset_by_node(full_ds.shuffle(), rank=0, world_size=world_size)
 
 
+@pytest.mark.skip(reason="require distributed torch")
 @pytest.mark.parametrize("streaming", [False, True])
 @require_torch
 @pytest.mark.skipif(os.name == "nt", reason="execute_subprocess_async doesn't support windows")
@@ -95,6 +96,7 @@
     execute_subprocess_async(cmd, env=os.environ.copy())
 
 
+@pytest.mark.skip(reason="require distributed torch")
 @pytest.mark.parametrize(
     "nproc_per_node, num_workers",
     [
--- a/tests/utils.py	2023-05-06 08:43:16.251987543 +0200
+++ b/tests/utils.py	2023-05-06 08:44:24.467952870 +0200
@@ -50,8 +50,8 @@
 # Audio
 require_sndfile = pytest.mark.skipif(
     # On Windows and OS X, soundfile installs sndfile
-    find_spec("soundfile") is None or version.parse(importlib.metadata.version("soundfile")) < version.parse("0.12.0"),
-    reason="test requires sndfile>=0.12.1: 'pip install \"soundfile>=0.12.1\"'; ",
+    True,
+    reason="test requires librosa",
 )
 
 # Beam
--- a/tests/features/test_audio.py	2023-05-06 09:03:58.680108142 +0200
+++ a/tests/features/test_audio.py	2023-05-06 09:05:50.463407967 +0200
@@ -57,6 +57,7 @@
     assert features.arrow_schema == pa.schema({"sequence_of_audios": pa.list_(Audio().pa_type)})
 
 
+@pytest.mark.skip(reason="require librosa")
 @pytest.mark.parametrize(
     "build_example",
     [
@@ -81,6 +82,7 @@
     assert decoded_example.keys() == {"path", "array", "sampling_rate"}
 
 
+@pytest.mark.skip(reason="require librosa")
 @pytest.mark.parametrize(
     "build_example",
     [
@@ -148,6 +149,7 @@
     assert decoded_example["sampling_rate"] == 48000
 
 
+@pytest.mark.skip(reason="require librosa")
 @pytest.mark.parametrize("sampling_rate", [16_000, 48_000])
 def test_audio_decode_example_pcm(shared_datadir, sampling_rate):
     audio_path = str(shared_datadir / "test_audio_16000.pcm")
@@ -414,6 +417,7 @@
     assert column[0]["sampling_rate"] == 16000
 
 
+@pytest.mark.skip(reason="require librosa")
 @pytest.mark.parametrize(
     "build_data",
     [
@@ -438,6 +442,7 @@
     assert item["audio"].keys() == {"path", "array", "sampling_rate"}
 
 
+@pytest.mark.skip(reason="require librosa")
 def test_dataset_concatenate_audio_features(shared_datadir):
     # we use a different data structure between 1 and 2 to make sure they are compatible with each other
     audio_path = str(shared_datadir / "test_audio_44100.wav")
@@ -451,6 +456,7 @@
     assert concatenated_dataset[1]["audio"]["array"].shape == dset2[0]["audio"]["array"].shape
 
 
+@pytest.mark.skip(reason="require librosa")
 def test_dataset_concatenate_nested_audio_features(shared_datadir):
     # we use a different data structure between 1 and 2 to make sure they are compatible with each other
     audio_path = str(shared_datadir / "test_audio_44100.wav")
@@ -610,6 +616,7 @@
     assert isinstance(ds, Dataset)
 
 
+@require_sndfile
 def test_dataset_with_audio_feature_undecoded(shared_datadir):
     audio_path = str(shared_datadir / "test_audio_44100.wav")
     data = {"audio": [audio_path]}
@@ -627,6 +634,7 @@
     assert column[0] == {"path": audio_path, "bytes": None}
 
 
+@require_sndfile
 def test_formatted_dataset_with_audio_feature_undecoded(shared_datadir):
     audio_path = str(shared_datadir / "test_audio_44100.wav")
     data = {"audio": [audio_path]}
@@ -658,6 +666,7 @@
         assert column[0] == {"path": audio_path, "bytes": None}
 
 
+@require_sndfile
 def test_dataset_with_audio_feature_map_undecoded(shared_datadir):
     audio_path = str(shared_datadir / "test_audio_44100.wav")
     data = {"audio": [audio_path]}
--- a/tests/packaged_modules/test_audiofolder.py	2023-05-06 14:00:39.560876163 +0200
+++ b/tests/packaged_modules/test_audiofolder.py	2023-05-06 14:01:26.005212423 +0200
@@ -1,10 +1,8 @@
 import shutil
 import textwrap
 
-import librosa
 import numpy as np
 import pytest
-import soundfile as sf
 
 from datasets import Audio, ClassLabel, Features, Value
 from datasets.data_files import DataFilesDict, get_data_patterns
@@ -192,8 +190,11 @@
     return data_files_with_two_splits_and_metadata
 
 
+@pytest.mark.skip(reason="require soundfile")
 @pytest.fixture
 def data_files_with_zip_archives(tmp_path, audio_file):
+    import soundfile as sf
+    import librosa
     data_dir = tmp_path / "audiofolder_data_dir_with_zip_archives"
     data_dir.mkdir(parents=True, exist_ok=True)
     archive_dir = data_dir / "archive"
--- a/tests/test_streaming_download_manager.py	2023-08-26 07:33:41.937389401 +0200
+++ b/tests/test_streaming_download_manager.py	2023-08-26 07:37:22.521218698 +0200
@@ -218,6 +218,7 @@
     assert output_path == _readd_double_slash_removed_by_path(Path(expected_path).as_posix())
 
 
+@pytest.mark.skip(reason="not working in sandbox")
 @pytest.mark.parametrize(
     "input_path, exists",
     [
@@ -301,6 +302,7 @@
         assert list(f) == TEST_URL_CONTENT.splitlines(keepends=True)
 
 
+@pytest.mark.skip(reason="not working in sandbox")
 @pytest.mark.parametrize(
     "input_path, expected_paths",
     [
@@ -331,6 +333,7 @@
         xlistdir(root_url, download_config=download_config)
 
 
+@pytest.mark.skip(reason="not working in sandbox")
 @pytest.mark.parametrize(
     "input_path, isdir",
     [
@@ -358,6 +361,7 @@
     assert xisdir(root_url, download_config=download_config) is False
 
 
+@pytest.mark.skip(reason="not working in sandbox")
 @pytest.mark.parametrize(
     "input_path, isfile",
     [
@@ -382,6 +386,7 @@
     assert xisfile(root_url + "qwertyuiop", download_config=download_config) is False
 
 
+@pytest.mark.skip(reason="not working in sandbox")
 @pytest.mark.parametrize(
     "input_path, size",
     [
@@ -407,6 +412,7 @@
         xgetsize(root_url + "qwertyuiop", download_config=download_config)
 
 
+@pytest.mark.skip(reason="not working in sandbox")
 @pytest.mark.parametrize(
     "input_path, expected_paths",
     [
@@ -450,6 +456,7 @@
     assert len(xglob("zip://qwertyuiop/*::" + root_url, download_config=download_config)) == 0
 
 
+@pytest.mark.skip(reason="not working in sandbox")
 @pytest.mark.parametrize(
     "input_path, expected_outputs",
     [
@@ -540,6 +547,7 @@
     def test_xpath_as_posix(self, input_path, expected_path):
         assert xPath(input_path).as_posix() == expected_path
 
+    @pytest.mark.skip(reason="not working in sandbox")
     @pytest.mark.parametrize(
         "input_path, exists",
         [
@@ -555,6 +563,7 @@
             (tmp_path / "file.txt").touch()
         assert xexists(input_path) is exists
 
+    @pytest.mark.skip(reason="not working in sandbox")
     @pytest.mark.parametrize(
         "input_path, pattern, expected_paths",
         [
@@ -593,6 +602,7 @@
         output_paths = sorted(xPath(input_path).glob(pattern))
         assert output_paths == expected_paths
 
+    @pytest.mark.skip(reason="not working in sandbox")
     @pytest.mark.parametrize(
         "input_path, pattern, expected_paths",
         [
--- a/tests/io/test_parquet.py	2024-02-22 19:19:53.890749240 +0100
+++ b/tests/io/test_parquet.py	2024-02-22 19:20:30.954099914 +0100
@@ -69,6 +69,7 @@
     _check_parquet_dataset(dataset, expected_features)
 
 
+@pytest.mark.skip()
 def test_parquet_read_geoparquet(geoparquet_path, tmp_path):
     cache_dir = tmp_path / "cache"
     dataset = ParquetDatasetReader(path_or_paths=geoparquet_path, cache_dir=cache_dir).read()